Grandparents’ and young adult grandchildren’s identity gaps and perceived caregiving intentions

Abstract

Utilizing communication theory of identity, this study examined identity gaps in relation to grandparents’ (GPs') and their young adult grandchildren's (GC's) communication satisfaction and, in turn, perceptions of GC’s caregiving intentions. Actor–partner interdependence models were examined with path analyses, using self-reported, cross-sectional survey data from GPs and GC (200 dyads). In regard to actor effects, as GPs and GC experienced identity gaps, they were less likely to report communication satisfaction. With lower levels of communication satisfaction, GPs were less likely to believe their GC would provide care for them in the future and GC had weaker intentions to provide care in the future. For partner effects, GP identity gaps were indirectly associated with grandchild caregiving intentions through either GP or GC communication satisfaction. By contrast, GC identity gaps were not significantly indirectly associated with GP perceptions of GC caregiving intentions through either GP or GC communication satisfaction.

Keywords

Actor–partner interdependence models caregiving grandchildren grandparents identity gaps

Recent statistics indicate that a growing number of adults 65 years or older require caregiving because of increased life spans and poor health care (Geurts, Van Tilburg, & Poortman, 2012; Wolff & Kasper, 2006). The U.S. Census Bureau estimated that the population of adults 65 years or older would increase from 40 million in 2010 to more than 88 million by 2050, eventually constituting 20% of the U.S. population (Vincent & Velkoff, 2010). The anticipated increase in older adults means a potential rise in the number of adults that require future care, yet the number of family members who are willing and able to provide assistance is unlikely to match this demand (Spillman & Pezzin, 2000). Although caregiving is often left to one adult child in the family (Merrill, 1996), research indicates that a substantial number of grandchildren (GC) assist with the care or become primary caregivers (Fruhauf, Jarrott, & Allen, 2006; Piercy & Chapman, 2001; Ross & Aday, 2006; Soliz, Lin, Anderson, & Harwood, 2006). Little is known, however, about motivational factors that lead GC to serve as primary caregivers for their aging grandparent (GP) or to assist a parent(s) in caring for their GP.

Consequently, the current study focuses on how GPs’ and GC's identities and communication satisfaction predict GC's intentions to care for GPs in the future. Examining the factors that predict GP-GC communication satisfaction is important because the quality of such communication may have implications for each individual’s relational decisions, such as future caregiving (Piercy & Chapman, 2001). For example, the willingness of young adult GC to assume caregiving responsibilities is important to consider because family collaboration is associated with decreased caregiver burden (Heru, Ryan, & Iqbal, 2004). In some cases, GC may reluctantly and halfheartedly care for their GPs, which may be attributed to the quality of GP-GC communication (Piercy & Chapman, 2001). By taking a dyadic approach to understanding identity and communication quality between GPs and their young adult GC, this study represents a first step in explaining GC's caregiving intentions and their GPs’ perceptions of such intentions.

Because GPs and GC belong to different age cohorts, the internalization and communication of their identities, as well as their expectations of each other’s identities as GPs or GC, are likely to vary. For example, GPs and GC may share a relational identity, but the complementary or conflicting nature of personal and relational identities may determine how satisfied they are with their communication and, in turn, perceptions of their GC's caregiving intentions (Kam & Hecht, 2009). What is more, the GP-GC relationship is involuntary (i.e., GC typically do not choose who their GPs are and vice versa), but maintaining the relationship requires deliberate involvement from both parties (Harwood & Lin, 2000; Mansson, Myers, & Turner, 2010). Such involvement may be motivated by each individual’s unique identity and the dyad’s communication quality. As such, an examination of the discrepancies in GPs’ and GC's experiences of personal and relational identities on the GP-GC relationship is warranted.

The communication theory of identity (CTI; Hecht, 1993) is an ideal lens to understand how inconsistencies between different frames of identity may contribute to the GP-GC relationship. CTI allows for an exploration of the lack of correspondence that sometimes occurs between young adult GC's self-perceptions and how they portray themselves (e.g., self-censorship; Williams & Giles, 1996). More specifically, Jung and Hecht (2004) posited that multiple frames of identity coexist, and when these frames oppose each other, identity gaps emerge. In the context of the GP-GC relationship, Kam and Hecht (2009) revealed that as young adult GC experienced identity gaps, they were less likely to feel satisfied with their communication. The current study extends this line of research by suggesting that as GPs and GC experience identity gaps, they are more likely to feel dissatisfied with their conversations. Subsequently, lower communication satisfaction may mean that GC are less likely to report strong caregiving intentions, and GPs may be less likely to believe their GC will provide future care. By contrast, more consistency between identities (i.e., smaller identity gaps) is likely related to greater communication satisfaction and, in turn, greater intentions to provide care in the future. With dyadic data, this study investigates these associations with respect to actor effects as well as how these associations manifest in regard to partner effects.

The emergence of identity gaps from CTI

Hecht’s (1993) CTI provides a valuable framework for understanding GP-GC relationships. CTI theorists conceptualized identity as individuals’ perceptions of self, which they enact, exchange, maintain, and alter in communication, relationships, and group memberships (Hecht, Warren, Jung, & Krieger, 2005). The section that follows explicates Hecht’s (1993) four identity frames, namely, personal, enacted, relational, and communal.

CTI’s four frames of identity

Individuals’ self-concept constitutes the personal frame of identity. For example, GPs perceiving themselves as kind, supportive, or wise are all aspects of the personal frame. The enacted frame is the expression or performance of identity through individuals’ communication (Jung & Hecht, 2008). Individuals may enact their identity as a GP by telling stories to their GC about their family history or providing emotional and financial support. The relational frame refers to individuals’ perceptions of themselves as part of relational units such as the GP-GC relationship. Individuals also internalize the other relational party’s perceptions and expectations of them (Jung & Hecht, 2004). If GPs believe their GC should assist them in their daily activities, then over time and through reinforcement, GC are likely to provide assistive care because they perceive that the GC’s role requires such. Hecht (1993) also suggested that individuals have a communal frame of identity, which refers to how a collective identity is ascribed by society. A commercial of a GP appearing fragile and walking slowly, which represents a stereotype of GPs and older adults in general, exemplifies the communal frame of identity.

Identity gaps

Hecht, Warren, Jung, and Krieger (2005) argued that frames do not exist in isolation but instead function in relative cohesion or opposition. To explicate the idea that frames of identity may operate inconsistently, Jung and his colleagues (e.g., Jung & Hecht, 2004, 2008; Jung, Hecht, & Wadsworth, 2007; Wadsworth, Hecht, & Jung, 2008) developed the concept of identity gaps. A number of gaps may exist among frames of identity, but this study concentrates on two that past research has investigated, namely, the personal–relational and personal–enacted gaps.

The personal–relational identity gap occurs when an individual’s self-concept does not match their perception of how a relational counterpart views him or her (Jung & Hecht, 2004). In particular, GPs may perceive themselves as strong and youthful (personal frame), yet believe their GC view them as fragile and old (relational frame), thus experiencing the personal–relational identity gap. Individuals experience the personal–enacted identity gap when they communicate in ways that are inconsistent with their self-concept (Jung & Hecht, 2008). More specifically, young adult GC may view themselves as an adult (personal frame) but behave like children (enacted frame) when interacting with their GP. The experience of different types of identity gaps may have negative outcomes for the involved parties (e.g., Drummond & Orbe, 2010; Faulkner & Hecht, 2011), which this study discusses in the following section.

Identity gaps in relation to communication satisfaction and caregiving intentions

When GPs and young adult GC interact, they may experience inconsistencies between their frames of identity, which in turn are likely to predict their communication satisfaction (Kam & Hecht, 2009). Communication satisfaction refers to the positive affect that individuals experience as they exchange messages that meet their expectations and internal standards (Hecht, 1978). Personal–relational and personal–enacted identity gaps may lead to lower levels of communication satisfaction because they can threaten GPs’ and GC's notions and portrayals of their “real” selves, which may lead them to communicate in ways that meet the other’s expectations and standards but not their own (Hecht & Faulkner, 2000; Williams & Nussbaum, 2001). For example, when GPs experience a personal–enacted gap, they may feel that they must behave in ways that portray a certain image but that is not true to their perceived personal identity. Feeling obligated to behave a certain way to meet their GC’s expectations of them may place pressure on the GP; if the GP’s expectation for satisfying communication is to be authentic to oneself, then a personal–enacted gap may lead to less satisfying communication. In addition, personal–relational and personal–enacted identity gaps may lead GPs and GC to communicate based on their misperceptions of each other’s real selves (e.g., engage in “elder speak” because of age-related stereotypes). As a result, such behavior may fail to meet the other person’s expectations for satisfying communication (Kam & Hecht, 2009; Soliz, 2007; Soliz & Hardwood, 2006). For example, when GPs experience a personal–relational gap, they may view themselves in a certain way but believe that their GC perceive them differently. Feeling that one’s GC do not know his or her real or authentic self means that when they interact, GC may behave in ways that do not meet the GP’s expectations or standards because the GC does not have a sense of who his or her GP really is as an individual.

Prior research has provided support for the negative relationships between identity gaps and communication satisfaction. College students felt less communication satisfaction with their classmates as they experienced personal–relational and personal–enacted identity gaps (Jung & Hecht, 2004). Within the context of the GP-GC relationship, Kam and Hecht (2009) found that as young adult GC experienced the personal–enacted gap, they were less satisfied with their communication with a GP. Such studies indicate a clear relationship between identity gaps and communication satisfaction. The current study extends this line of research by investigating how identity gaps are related to GC's caregiving intentions and their GPs’ perceptions of such intentions through communication satisfaction.

Examining the factors that predict GP-GC communication satisfaction is important because the quality of such communication may have implications for each individual’s relational decisions, such as future caregiving. Understanding GC's intentions to care for their GPs in the future has serious long-term implications for the growing number of older adults in the U.S., who require care (Fruhauf et al., 2006). For example, as GC have smaller identity gaps, they may be more likely to enjoy higher levels of communication satisfaction and, in turn, may be more likely to voluntarily provide care for their GPs. Indeed, Hamill (2012) found that young adult GC who showed affection toward GPs suffering from Alzheimer’s disease provided greater assistance to their ailing GPs, which suggests that GP-GC communication quality is important when considering caregiving intentions. By contrast, the less GPs and GC derive satisfaction from their conversations, the less likely GPs are to anticipate that their GC will provide care if they require such assistance. In addition, the less likely GC are to form caregiving intentions. Although Piercy and Chapman (2001) did not specifically consider communication satisfaction, one granddaughter in their interview study reported feeling reluctant and avoidant in contributing to her grandmother’s care because of a past experience when the grandmother rejected her. Taken together, we can surmise that identity gaps and communication satisfaction between GPs and GC may predict GC's involvement in the caregiving process. Thus, it was hypothesized that:

H1: Communication satisfaction mediates the associations between identity gaps and perceived caregiving intentions such that (a) identity gaps are negatively associated with communication satisfaction and (b) communication satisfaction is negatively associated with perceived caregiving intentions.

The interdependent associations between GPs and GC

Thus far, this study has focused on individual-level associations in which GPs’ identity gaps are indirectly related to their perceptions of their GC's caregiving intentions through GPs’ communication satisfaction. Similarly, GC's identity gaps are indirectly related to their caregiving intentions through GC's communication satisfaction. These individual-level associations reflect actor effects, which refer to the individual’s scores on independent variables in relation to their scores on dependent variables (Ramirez, 2008). To date, research on identity gaps (e.g., Jung, 2013) and GP-GC communication (e.g., Hamill, 2012) has primarily focused on how one member experiences a particular communication, relational, or psychological phenomenon. Yet the perceptions and behaviors of both members of the GP-GC dyad are likely to be related to each other (Kemp, 2007; Nussbaum & Bettini, 1994) and these partner associations should also be examined.

Although examining actor effects is informative, it is equally important to consider partner effects, given that GP-GC identity gaps and communication are likely to be interdependent. Several prior studies have highlighted the interdependent nature of the GP-GC relationship. For example, as GC reported closer GP-GC relationships, they were more likely to spend time with GPs and, in turn, more likely to be influenced by GPs’ beliefs and values (Brussoni & Boon, 1998). Nevertheless, missing from past literature on the interdependent nature of GPs and GC is an understanding of how the identities of both parties relate to their communication satisfaction and, in turn, perceptions of future caregiving. Soliz and Harwood (2006) articulated the interdependent nature of GPs’ and GC's communication as shared family identity from the GC's perspective, but the current study focuses on the communication of identity specific to the dyad.

By investigating the partner effects of identity gaps on communication satisfaction (and indirectly on future caregiving intentions), this study may shed light on how a satisfying GP-GC relationship may be related to greater intentions for providing future care. As such, this study advances our understanding of identity in the GP-GC relationship by considering the interdependence between members of the intergenerational relationship. The GP-GC relationship can only completely be understood through a broader relational approach that “emphasizes the individual’s place within a dynamic and continuous set of transactional processes” (Hillcoat-Nallétamby & Phillips, 2011, p. 212). Thus, both the GP and the GC must be seen within an interactional network where both generations have agency over the nature of the relationship.

Past studies (e.g., Villar, Triado, Pinazo-Hernandis, Celdran, & Solé, 2010) emphasized the importance of comparing the perspectives of both GPs and GC to better understand intergenerational relationships inside the family. For example, Matos and Neves (2012) found that GC's communication could alter their GPs’ dietary behavior and attitudes toward contemporary social issues. In this study, we propose both partner and actor effects for GPs’ identity gaps in relation to communication satisfaction and future caregiving intentions for both members of the dyad. The interdependent nature of the GP-GC relationship suggests that GPs experiencing identity gaps may also be related to GC's communication satisfaction, particularly if GC perceive their GPs are not acting authentically. In turn, if GC are not satisfied with their communication, they may be less invested in the relationship and less inclined to provide care in the future. Furthermore, if GPs are not satisfied with their communication because they experience identity gaps, GC may perceive that their GP has low communication satisfaction. In turn, GC may be less likely to provide care in the future because they may similarly derive less satisfaction from their interactions. Hence, the following hypothesis is set forth:

H2: GPs’ identity gaps are indirectly related to GC's caregiving intentions such that:

GP identity gaps are negatively related to GP communication satisfaction, and GP communication satisfaction is positively related to GC caregiving intentions.

GP identity gaps are negatively related to GC communication satisfaction, and GC communication satisfaction is positively related to GC caregiving intentions.

In a dyadic analysis of GPs and GC, Villar, Triado, Pinazo-Hernandis, Celdran, and Solé (2010) found that GPs perceived a greater level of change in the GP-GC relationship as GC age and that older GC had less optimistic perceptions of the relationship. Because young adult GC may experience several shifts in their own identity during the transition to adulthood (Arnett, 2000), GPs’ perceptions of caregiving intentions may be adversely affected. Thus, we propose that, as GC experience identity gaps, they are less likely to feel satisfied with their conversations, and in recognizing such dissatisfaction, their GPs are less likely to believe their GC have intentions to care for them in the future. Similarly, as GC experience identity gaps, GPs are less likely to feel satisfied with the conversations and, in turn, GPs are less likely to believe their GC would care for them in the future. For these reasons, the following hypothesis was posited:

H3: GC identity gaps are indirectly related to GPs’ perceptions of their GC’s caregiving intentions such that:

GC identity gaps are negatively related to GC communication satisfaction, and GC communication satisfaction is positively related to GPs’ perceptions of their GC's caregiving intentions.

GC identity gaps are negatively related to GP communication satisfaction, and GP communication satisfaction is positively related to GPs’ perceptions of their GC's caregiving intentions.

Method

Participants

This study is based on self-reported, cross-sectional survey data provided by 200 GP-GC dyads. The mean age was 74.42 years (SD = 7.62; 60–94 years) for GPs and 19.86 years (SD = 1.69; 18–38 years) for young adult GC. Among the GPs, 68% were female and 32% were male, whereas 33% of the young adult GC were female and 68% were male. The sample comprised of mostly non-Latino White participants (94%), with 3% Asian or Pacific Islander, 1% African American/Black, .5% Latino, and 1.5% other (e.g., of mixed ethnicity). Among GPs, 14% were paternal grandfathers, 25% paternal grandmothers, 18% maternal grandfathers, and 43% maternal grandmothers.

Procedures

Undergraduate students who were enrolled in an introductory communication course were asked to complete an online survey. To participate, students had to be able to recall their relationship with one of their living biological GPs who had Internet access. Students were allowed to choose any GP as long as he/she fit these criteria. In addition, students were asked to contact the GP who they reflected on and ask this GP to complete a similar online survey. Students were provided with a description of the study, a message for the GP requesting his/her participation, and a link to the survey, which students were asked to relay to their GP. In the study description, students were told that they would receive research credit for completing their online survey regardless of whether their GP completed it. Students were discouraged from completing the GP survey themselves because they would receive credit despite a GP’s nonparticipation.

To match the responses, students created an identification code that they typed into their online survey and then shared the code with their GP who typed the code into his/her own survey. Neither member had access to the other’s responses. GPs and young adult GC were asked to complete the survey alone and to keep their responses private. The survey took approximately 30–45 min to complete.

Measures

GPs and young adult GC used a 5-point scale (1 = strongly disagree to 5 = strongly agree) to respond to each measure for all the variables. To assess the measures, confirmatory factor analyses (CFAs) were conducted in Mplus (Muthén & Muthén, 1998–2007). A series of CFAs were examined, and items were dropped one by one if their standardized factor loadings were <.60 and the model fit was unacceptable. A well-fitting model should have a root mean square error of approximation (RMSEA) ≤.06 (Hu & Bentler, 1999), but a reasonably fitting model can have an RMSEA of <.08 (Browne & Cudeck, 1993). Further, for a well-fitting model, the comparative fit index (CFI) should be ≥.95; however, an acceptably fitting model can have a CFI value ≥.90 (Beaudoin & Thorson, 2006; Hu & Bentler, 1999). Lastly, the standardized root mean square residual (SRMR) should be <.08 (Hu & Bentler, 1999).

This study’s limited sample size of 200 dyads did not allow for the examination of an omnibus CFA model that included all the latent factors and their corresponding indicators (i.e., two identity gaps, communication satisfaction, and caregiving intentions). Thus, each scale was inspected with separate CFAs (e.g., only examined communication satisfaction by itself). For each CFA, both GPs’ and young adult GC's responses were included in the same model. In addition, the GPs’ and young adult GC's latent factors were correlated as well as their corresponding error terms (Kenny, Kashy, & Cook, 2006, pp. 106–108). The factor loadings were also constrained to be equal across GPs and young adult GC, as recommended by Kenny et al. For these reasons, one CFA result is reported for each measure.

Identity gaps

GPs and young adult GC completed a modified version of the Identity Gap Scale (Jung & Hecht, 2004) in which they reported their personal–relational and personal–enacted gaps. A shortened version of the personal–relational scale was used, which consisted of 5 items (RMSEA = .05, 90% confidence interval (CI) = [.00, .08]; CFI = .97; SRMR = .06). From the GPs’ perspective, the 5 items are (1) “I agree with how my grandchild would describe me,” (2) “I am different from the way my grandchild would see me,” (3) “I feel that my grandchild sees me as I see myself,” (4) “I feel that my grandchild has correct information about me,” and (5) “I feel there is no difference between who I think I am and who my grandchild thinks I am.” Items 1, 3, 4, and 5 were reverse coded.

Similar to the personal–relational gap, a shortened version of the Personal–Enacted Scale was used, which consisted of 9 items (RMSEA = .07, 90% CI = [.06, .08]; CFI = .88; SRMR = .07). From the GPs’ perspective, the 9 items are (1) “When I communicate with my grandchild, he/she gets to know the real me,” (2) “I express myself in a certain way that is not the real me when communicating with my grandchild,” (3) “I do not reveal important aspects of myself in communication with my grandchild,” (4) “I do not express the real me when, I think, it is different from my grandchild’s expectation,” (5) “I sometimes mislead my grandchild about who I really am,” (6) “I speak truthfully to my grandchild about myself,” (7) “I freely express real me in communication with my grandchild,” (8) “I feel that I can be myself when communicating with my grandchild,” and (9) “I feel that I can communicate with my grandchild in a way that is consistent with who I really am.” Items 1, 6, 7, 8, and 9 were reverse coded. Afterward, a CFA model with both identity gaps and their corresponding items (for both GP and GC) was examined, and the model fits the data adequately (RMSEA = .07, 90% CI = [.06, .07]; CFI = .86; SRMR = .07), although the CFI value was low.

Communication satisfaction

GPs and young adult GC completed the 5-item Interpersonal Communication Satisfaction Inventory (Hecht, 1978) that was adapted by Lin and Harwood (2003). Participants were asked to report on the “conversations [they] typically have with the GP.” Of the 5 items, GP sample items include “I am satisfied with my conversations with my grandchild” or “Our conversations flow smoothly” (RMSEA = .07, 90% CI = [.04, .09]; CFI = .92; SRMR = .08).

Perceptions of GC's caregiving intentions

GPs and young adult GC completed a shortened modified version of the Filial Responsibility Scale (Hamon, 1988). As an example, the modified introduction statement for GPs was:

There are many ways in which GC handle their grandparents’ needs, so different GC react in different ways. In turn, grandparents have different expectations for their GC. Furthermore, other family members, such as your son or daughter or your spouse may primarily assist you. Thus, the following questions ask how involved you think your grandchild would be in caring for you.

The scale comprised of 11 items, which are listed as follows from the GP perspective: (1) “act as one of my caregivers,” (2) “give me financial support,” (3) “make room in his/her home for me to live,” (4) “sacrifice personal freedom to care for me,” (5) “feel responsible for me,” (6) “live with me,” (7) “visit once a week,” (8) “communicate with me once a week,” (9) “adjust his/her schedule to help me,” (10) “live close to me,” and (11) “adjust his/her work schedule to help me” (RMSEA = .07, 90% CI = [.06, .08]; CFI = .92; SRMR = .07).

Control variables

Preliminary analyses revealed that GPs’ and GC's gender was not significantly related to the dependent variables; therefore, it was excluded as a control variable. GP type, GPs’ and young adult GC's reports of the most commonly used method of communicating and communication frequency were significantly related to several of the dependent variables; therefore, they were included as control variables. Paths were drawn from these variables to the corresponding dependent variables in which they shared significant associations. Forty-six percent of the GPs reported communicating with their GC a few times a month, 39.4% a few times a year, none reported once or less than once a year, and 13.6% almost never. More specifically, 53.3% of the GC reported communicating with their GP a few times a month, 35.2% a few times a year, and 11.6% almost never. Further, 46.9% of the GPs reported face to face as the most commonly used way of communicating with their GC, followed by the telephone (30.1%), e-mail (13.3%), through a family member (4.6%), multiple media (3.6%), text messaging (1%), and letters through the mail (.5%). For GC, 46.5% most often communicated via face to face, telephone (34.0%), e-mail (16.5%), and a family member (3%).

Preliminary analyses

The items that formed each scale were averaged to create composite scores (see Table 1 for bivariate correlations, descriptive statistics, and reliability coefficients). Afterward, paired-sample t-tests were conducted to determine whether GPs and young adult GC significantly differed from each other in their mean identity gaps, communication satisfaction, and caregiving intentions. As seen in Table 1, GC, as compared to GPs, reported significantly higher means for the personal–relational gap, t(187) = 4.76, p < .001; M = 2.42, SD = .70 vs. M = 2.15, SD = .62; Cohen’s d = .35, personal–enacted gap, t(182) = 7.28, p < .001; M = 2.26, SD = .63 vs. M = 1.87, SD = .55; Cohen’s d = .54, and caregiving intentions, t(180) = 5.22, p < .001; M = 3.71, SD = .71 vs. M = 3.41, SD = .76; Cohen’s d = .39. Nevertheless, GPs reported significantly greater communication satisfaction than GC, t(188) = −5.75, p < .001; M = 4.16, SD = .55 vs. M = 3.90, SD = .59; Cohen’s d = .42. At the bivariate level, GPs’ and young adult GC’s identity gaps were positively associated with each other but negatively associated with communication satisfaction and caregiving intentions (see Table 1).

Table 1.

Bivariate correlations and reliabilities for GPs and young adult GC.

	Means (SD)	1	2	3	4	5	6	7	8
1. GP personal–relational gap	2.16 (.62)	.81
2. GP personal–enacted gap	1.89 (.58)	.59**	.88
3. GP communication satisfaction	4.16 (.55)	−.54**	−.66**	.70
4. GP perceptions of GC caregiving intentions	3.41 (.76)	−.27**	−.26**	.29**	.94
5. GC personal–relational gap	2.40 (.69)	.32**	.25**	−.25**	−.15*	.85
6. GC personal–enacted gap	2.26 (.62)	.25**	.23**	−.28**	−.11	.70**	.89
7. GC communication satisfaction	3.91 (.58)	−.30**	−.32**	.41**	.13	−.60**	−.64**	.71
8. GC caregiving intentions	3.73 (.70)	−.18*	−.19**	.29**	.45**	−.43**	−.37**	.39**	.93

Note. The diagonal contains Cronbach’s α coefficients. All constructs were operationalized based on a 5-point scale (1 = strongly disagree to 5 = strongly agree). GP = grandparent and GC = grandchild.

*p < .05; **p < .01.

The intraclass correlations (ICCs) were obtained to determine the extent to which variability could be attributed to within-person or between-person variability. The ICC values were .87 for the personal–relational gap, .86 for the personal–enacted gap, .95 for communication satisfaction, and .91 for caregiving intentions. With the values closer to 1.0, the variability can primarily be attributed to between-person variability.

Finally, although GPs and young adult GC are theoretically distinguishable, a formal test was conducted. Two separate models were examined, one for each identity gap. In each model, the actor effects, partner effects, error variances, Y intercepts, X variances, and X means were constrained to be equal based on GP or GC status. The tests of complete indistinguishability were all significant (p < .001), thereby indicating that the dyads should be treated as distinguishable (Kenny et al., 2006).

Substantive analyses

To test the hypotheses, two actor–partner interdependence mediation models (Kenny et al., 2006) were examined using Mplus. Each model included one of the two identity gaps for both GPs and young adult GC. Structural equation modeling with observed variables (i.e., path analysis) was applied, using the full information maximum likelihood (missingness for each variable was less than 5%) with robust standard errors to handle the data’s nonnormality. Finally, to test for mediation, direct and indirect paths were simultaneously examined, and as suggested by Preacher and Hayes (2008), bootstrap bias-corrected CIs were obtained, with the number of bootstrapping set to 5,000.

Results

The model with the personal–relational gap (see Figure 1) fits the data acceptably, χ²(32) = 53.87, p < .01; CFI = .93; RMSEA = 0.06, 90% CI = [.03, .09]; SRMR = .08, as did the model with the personal–enacted gap, χ²(32) = 59.92, p < .01; CFI = .92; RMSEA = 0.07, 90% CI = [.04, .09]; SRMR = .08 (see Figure 2). The following sections describe the detailed results for the hypotheses.

Figure 1.

An actor–partner interdependence mediation model for grandparents and young adult GC with personal–relational identity gaps.

Figure 2.

An actor–partner interdependence mediation model for grandparents and young adult GC with personal–enacted identity gaps.

H1: Actor effects on caregiving intentions by way of indirect associations

H1 posited actor effects such that communication satisfaction would mediate the associations between identity gaps and caregiving intentions for GPs and for young adult GC (see Table 2 for indirect associations). The following paragraphs describe the individual paths and the indirect associations that correspond to this hypothesis. As suggested by Slater, Hayes, and Ford (2007), because this study examined mediation, the unstandardized results are reported with respect to the indirect associations and the figures.

Table 2.

Actor and partner effects on caregiving intentions by way of indirect effects.

H1: Model with the personal–relational identity gap	Indirect effects (95% CIs)
1. GP personal–relational gap → GP communication satisfaction → GP perceptions of GC’s caregiving intentions	−.12 (−.282, −.040)*
2. GC personal–relational gap → GC communication satisfaction → GC caregiving intentions	−.08 (−.181, −.012)*
H1: Model with the personal–enacted identity gap
1. GP personal–enacted gap → GP communication satisfaction → GP perceptions of GC’s caregiving intentions	−.144 (−.359, .046)
2. GC personal–enacted gap → GC communication satisfaction → GC caregiving intentions	−.12 (−.259, −.015)*
H2a: GP identity gaps → GP communication satisfaction → GC caregiving intentions
1. GP personal–relational gap → GP communication satisfaction → GC caregiving intentions	−.09 (−.195, −.022)*
2. GP personal–enacted gap → GP communication satisfaction → GC caregiving intentions	−.13 (−.270, −.010)*
H2b: GP identity gaps → GC communication satisfaction → GC caregiving intentions
1. GP personal–relational gap → GC communication satisfaction → GC caregiving intentions	−.02 (−.064, −.001)*
2. GP personal–enacted gap→ GC communication satisfaction → GC caregiving intentions	−.04 (−.106, −.007)*
H3a: GC identity gaps → GC communication satisfaction → GP perceptions of GC caregiving intentions
1. GC personal–relational gap → GC communication satisfaction → GP perceptions of GC caregiving intent	−.06 (−.062, .166)
2. GC personal–enacted gap → GC communication satisfaction → GP perceptions of GC caregiving-intent	.03 (−.123, .154)
H3b: GC identity gaps → GP communication satisfaction → GP perceptions of GC caregiving intentions
1. GC personal–relational gap→ GP communication satisfaction → GP perceptions of GC caregiving intent	−.02 (−.066, .002)
2. GC personal–enacted gap→ GP communication satisfaction → GP perceptions of GC caregiving intent	−.03 (−.084, .002)

Note. The asterisks indicate significant indirect effects (i.e., zero is not within the 95% confidence interval). GP = grandparent; GC = grandchild.

Personal–relational identity gap

For GPs, the personal–relational identity gap was negatively related to their communication satisfaction (β = −.50, b = −.44, t = −5.28, p < .001), and communication satisfaction was positively related to the perception that their GC would care for them in the future, if needed (β = .20, b = .28, t = 2.24, p < .001). The bootstrap 95% CI revealed that this indirect association was significant. GP personal–relational gap did not exhibit a significant direct effect (β = −.16, b = −.19, t = −1.78, p = .08), but it had a significant total effect (β = .25, b = −.30, t = −3.23, p < .001) on GP perception of GC’s caregiving intentions. For GC, the personal–relational identity gap was negatively related to their communication satisfaction (β = −.56, b = −.47, t = −9.18, p < .001), and communication satisfaction was marginally positively related to their caregiving intentions (β = .14, b = .17, t = 1.82, p = .07). The indirect association was significant. Moreover, the GC personal–relational gap had a significant direct effect (β = −.28, b = −.28, t = −3.42, p < .01) on GC caregiving intentions and a significant total effect (β = −.38, b = −.38, t = −4.69, p < .001).

Personal–enacted identity gap

For GPs, the personal–enacted identity gap was negatively related to their communication satisfaction (β = −.65, b = −.62, t = −10.56, p < .001), but communication satisfaction was not significantly related to the perception that their GC would care for them in the future, if needed (β = .17, b = .23, t = 1.41, p = .16). This indirect association was not significant. GP personal–enacted gap did not exhibit a significant direct effect (β = −.13, b = −.17, t = −1.19, p = .23), but it had a significant total effect (β = −.23, b = −.30, t = −3.11, p < .01) on GP perception of GC’s caregiving intentions. For GC, the personal–enacted identity gap was negatively related to their communication satisfaction (β = −.59, b = −.55, t = −8.48, p < .001) and communication satisfaction was positively related to their caregiving intentions (β = .18, b = .22, t = 2.08, p < .05). This indirect association was significant. The GC personal–enacted gap had a marginally significant direct effect (β = −.16, b = −.18, t = −1.88, p = .06) on GC caregiving intentions and a significant total effect (β = −.29, b = −.33, t = −3.71, p < .001).

A summary of H1

Across the two identity gaps, the directions of the associations were consistent with H1. Generally, identity gaps were negatively associated with communication satisfaction, and communication satisfaction was positively associated with caregiving intentions. Of the four possible indirect associations, three of them were significant. Thus, H1 was substantially supported.

H2: Partner effects on GC caregiving intentions by way of indirect associations

H2a posited that GP identity gaps would be negatively related to GP communication satisfaction, and GP communication satisfaction would be positively related to GC caregiving intentions (see Table 2 for indirect associations). GP personal–relational gap was negatively related to GP communication satisfaction, β = −.50, b = −.44, t = −5.28, p < .001, and GP communication satisfaction was positively related to GC caregiving intentions, β = .17, b = .21, t = 2.46, p < .05. This indirect association was significant. The GP personal–relational gap did not exhibit a significant direct effect, β = .07, b = .07, t = .841, p = .40, or a significant total effect, β = −.04, b = −.04, t = −.533, p = .59, on GC caregiving intention. GP personal–enacted gap was negatively related to GP communication satisfaction, β = −.65, b = −.62, t = −10.56, p < .001, and GP communication satisfaction was positively related to GC caregiving intentions, β = .17, b = .21, t = 1.98, p < .05. This indirect association was significant. The GP personal–enacted gap did not exhibit a significant direct effect, β = .04, b = .05, t = .545, p = .59, or a significant total effect, β = −.10, b = −.12, t = −1.53, p = .13, on GC caregiving intention.

H2b posited that GP identity gaps would be negatively related to GC communication satisfaction, and GC communication satisfaction would be positively related to GC caregiving intentions (see Table 2). GP personal–relational gap was negatively related to GC communication satisfaction, β = −.13, b = −.12, t = −2.12, p < .05, and GC communication satisfaction was marginally positively related to GC caregiving intentions, β = .14, b = .17, t = 1.82, p = .07. This indirect association was significant. GP personal–enacted gap was negatively related to GC communication satisfaction, β = −.19, b = −.19, t = −2.88, p < .01, and GC communication satisfaction was positively related to GC caregiving intentions, β = .18, b = .22, t = 2.08, p < .05. This indirect association was significant.

A summary of H2

Across the two identity gaps, the directions of the associations were consistent with H2. All four possible indirect associations were significant. GP identity gaps were indirectly associated with GC caregiving intentions through either GP or GC communication satisfaction. Thus, H2 was supported.

H3: Partner effects on GP perceptions of GC’s caregiving intentions

H3a posited that GC identity gaps would be negatively related to GC communication satisfaction, and GC communication satisfaction would be positively related to GP perceptions of his/her GC’s caregiving intentions (see Table 2 for indirect effects). GC personal–relational gap was negatively related to GC communication satisfaction, β = −.56, b = −.47, t = −9.18, p < .001, but GC communication satisfaction was not significantly related to GP perceptions of his/her GC’s caregiving intentions, β = −.09, b = −.12, t = −.979, p = .33. This indirect association was not significant. The GC personal–relational gap did not exhibit a significant direct effect, β = −.03, b = −.04, t = −.344, p = .73, or a significant total effect, β = .01, b = .01, t = .005, p = 1.0, on GP perception of GC’s caregiving intentions.

GC personal–enacted gap was negatively related to GC communication satisfaction, β = −.59, b = −.55, t = −8.48, p < .001, but GC communication satisfaction was not significantly related to GP perceptions of his/her GC’s caregiving intentions, β = −.04, b = −.05, t = −.412, p = .68. This indirect association was not significant. GC personal–enacted gap did not exhibit a significant direct effect, β = .03, b = .04, t = .315, p = .75, or a significant total effect, β = .04, b = .04, t = .433, p = .67, on GP perception of GC’s caregiving intentions.

H3b posited that GC identity gaps would be negatively related to GP communication satisfaction, and GP communication satisfaction would be positively related to GP perceptions of his/her GC’s caregiving intentions (see Table 2 for indirect effects). GC personal–relational gap was not significantly related to GP communication satisfaction, β = −.09, b = −.07, t = −1.57, p = .12, but GP communication satisfaction was positively related to GP perceptions of his/her GC’s caregiving intentions, β = .20, b = .28, t = 2.24, p < .05. This indirect association was not significant. GC personal–enacted gap was negatively related to GP communication satisfaction, β = −.12, b = −.11, t = −2.16, p < .05, but GP communication satisfaction was not significantly related to GP perceptions of his/her GC’s caregiving intentions, β = .17, b = .23, t = 1.41, p = .16. This indirect association was not significant.

A summary of H3

Across the two identity gaps, the directions of the associations were consistent with H3. Nevertheless, all four possible indirect associations were not significant. GC identity gaps were not significantly indirectly associated with GP perceptions of GC caregiving intentions through either GP or GC communication satisfaction. Thus, H3 was not supported.

Discussion

This study incorporated the CTI (Hecht, 1993) and identity gaps (Jung & Hecht, 2004) to understand GP-GC communication satisfaction and perceptions of caregiving intentions. The GP-GC relationship remains an understudied dyad, with GPs being members of a particular age-group that receives little research attention within the field of communication (Soliz et al., 2006). This study extends past research on the GP-GC relationship by representing a more complex process in which identity gaps may operate (via mediation) and by demonstrating the applicability of identity gaps to GPs (in addition to GC). The actor effects set forth in H1 were supported, except with respect to the GP personal–enacted gap. Generally, however, GPs and young adult GC experienced inconsistencies between their frames of identity that were associated with less satisfying conversations and, in turn, weaker intentions (or perceptions of such intentions) to provide care if needed. Furthermore, the partner effects posited in H2 garnered support, such that GP identity gaps were indirectly related to GC caregiving intentions through GP communication satisfaction or GC communication satisfaction. By contrast, the partner effects posited in H3 were not supported. GC identity gaps were not indirectly associated with GP perceptions of GC caregiving intentions through GC or GP communication satisfaction. The implications of these results are discussed below.

H1: Actor effects

H1 predicted that communication satisfaction would mediate the associations between identity gaps and perceived caregiving intentions such that (a) identity gaps are negatively associated with communication satisfaction and (b) communication satisfaction is negatively associated with perceived caregiving intentions. This study’s results primarily supported H1. As GPs and young adult GC experienced the personal–relational gap, they were less likely to feel satisfied with their communication and, in turn, less likely to report caregiving intentions or the belief in their GC's caregiving intentions. This indirect association was also apparent for GC as they experienced the personal–enacted gap but not for GPs.

Such findings support the notion that as GC report smaller gaps, they are likely to enjoy higher levels of communication satisfaction and, in turn, may be more likely to voluntarily provide care for their GPs. Similarly, as GPs experience smaller gaps, they are likely to report higher levels of communication satisfaction and, in turn, believe that their GC would provide care if she/he needed such assistance. Our findings are consistent with past scholarship, suggesting a direct relationship between individuals’ identity gaps and their communication satisfaction (e.g., Jung & Hecht, 2004; Kam & Hecht, 2009). This study extends past research on GC and identity gaps by showing how inconsistencies between frames of identity may have implications for young adult GC’s caregiving intentions and, more specifically, for the well-being of GPs. Taken together, our results suggest that GPs’ and young adult GC's identities may contribute to their communication quality and caregiving intentions (or perceptions of such intentions).

Nevertheless, one finding that warrants explication is the nonsignificant mediation of communication satisfaction for GPs experiencing a personal–enacted gap. Similar to the other identity gap, the personal–enacted gap was negatively related to communication satisfaction. Thus, at the root of the nonsignificant mediation is the nonsignificant association between GP communication satisfaction and GP perceptions of GC caregiving intentions, but this nonsignificant association was only found when taking into account the direct path from GP personal–enacted gap and GP perceptions of GC caregiving intentions. One explanation for this nonsignificant mediation is that, for GPs experiencing a personal–enacted gap, factors other than communication satisfaction play a role in their perceptions of GC's caregiving intentions. Thus, rather than focusing on communication satisfaction, it may be more fruitful to examine communication behaviors that occur when GPs communicate in ways that are inconsistent with their self-concept. For example, uncertainty about how to respond to disclosures (Williams & Nussbaum, 2001) or a low degree of communication competence (Fowler & Soliz, 2010) may help explain the relationship between GPs’ personal–enacted identity gap and perceptions of GC's caregiving intentions.

H2: Partner effects for GPs’ identity gaps

H2 posited partner effects such that GPs’ identity gaps would be indirectly related to GC's caregiving intentions through GP and GC communication satisfaction. The results yielded support for H2 concerning GPs’ identity gaps in relation to GC's caregiving intentions. First, as GPs experienced any of the two identity gaps, they were less likely to feel satisfied with their communication and, in turn, GC reported weaker caregiving intentions. In addition, as GPs experienced any of the two identity gaps, GC were less likely to be satisfied with their communication and, in turn, GC reported weaker caregiving intentions.

Our findings reveal that GPs’ identity gaps are related to GC's communication satisfaction and caregiving intentions, thereby providing evidence for partner effects. These results are in accordance with the large body of literature indicating the interdependence between GP and GC thoughts and behaviors. For example, family systems theory suggests that individuals cannot be understood in isolation from another; families are considered systems of interconnected and interdependent individuals, none of whom can be understood in isolation from the family system (Whitechurch & Constantine, 1993). Despite this notion, most of the extant literature on young adult caregivers focuses solely on the perspective of either GP or GC, which neglects the bidirectional flow of influence. GC, however, can influence a GP's behavior just as readily as a GP can influence their GC (Matos & Neves, 2012).

H3: Partner effects for GC's identity gaps

H3 set forth partner effects such that GC's identity gaps would be indirectly related to GPs’ perceptions of their GC's caregiving intentions through GC and GP communication satisfaction. Although our hypothesized partner effects were supported for H2, interestingly, H3 did not receive support. Young adult GC reported lower communication satisfaction when they experienced any of the two identity gaps; however, only the GC personal–enacted gap was significantly related to GP communication satisfaction. The positive association between GC identity gaps and GC communication satisfaction falls in line with past research; Kam and Hecht (2009) found that as young adult GC experienced identity gaps, they felt less satisfied with their communication involving a GP. Nevertheless, when extending such research by incorporating a dyadic perspective, this study found that only the GC's personal–enacted gap was negatively related to GPs’ communication satisfaction.

When including both GPs’ and GC's perspectives, the current investigation found only partial evidence for partner effects with respect to GC's identity gaps in relation to GPs’ communication satisfaction. Because the personal–relational gap represents a cognitive disagreement between frames of identity (Jung & Hecht, 2004), GPs’ communication satisfaction may have remained unaffected because they were simply unable to sense their GC's identity gap. In contrast, the significant negative relationship between GC's personal–enacted gap and GPs’ communication satisfaction is logical because this gap focuses on a behavioral conflict between frames of identity, for example, times when GC communicate with a GP in ways that are inconsistent with their own self-concept (Jung & Hecht, 2004). By identifying the outcomes of GC's personal–enacted identity gap for GPs, our results extend the notion that young adults portray inauthentic selves when interacting with older adults (Williams & Nussbaum, 2001). Interestingly, however, it appears that GPs may be more sensitive to their GC acting in ways that are inconsistent with their identity.

A final point worth noting is that GC's communication satisfaction was not significantly related to GPs’ perceptions of their GC’s caregiving intentions. One possible explanation for this nonsignificant finding is that GPs may not place weight on their GC's communication satisfaction as an indicator of their GC's future caregiving. Instead, GPs may base their GC's caregiving intentions on other factors such as affection for their GP, felt obligations to the GP, the desire to alleviate pressure for their parent(s), moral and religious beliefs, and the absence of other family members (Fruhauf et al., 2006; Piercy, 2007). For example, the young adult GC in Dellmann-Jenkins, Blankemeyer, and Pinkard’s (2000) study reported two common reasons for becoming caregivers, namely, filial duty and wanting to assist their parents. Furthermore, GPs may perceive GC's caregiving intentions as a stable trait rather than one that is contingent on the extent to which their GC is satisfied with their communication.

Theoretical implications

The findings presented in this study illustrate the significant role that GPs’ and GC's experience of identity gaps have in predicting communication satisfaction and future caregiving. The actor effects (H1) reported herein are consistent with previous research on identity gaps and communication satisfaction (e.g., Kam & Hecht, 2009). Likewise, the differential associations that the personal–enacted identity gap shared with communication satisfaction and future caregiving compared with the personal–relational identity gap is similar to previous studies (e.g., Jung et al., 2007; Kam & Hecht, 2009). It is noteworthy that GPs’ experiences with identity gaps revealed significant partner effects (H2), whereas GC's identity gaps did not yield significant results (H3). This finding highlights the importance of GPs’ identity, relative to GC's identity, in the GP-GC relationship and implies that identity gaps do function differently in this intergenerational relationship.

This study also has significant implications for the role that identity gaps play in communication outcomes and future caregiving, revealing an important theoretical development for the CTI (Hecht, 1993). Earlier tests of CTI have revealed associations between identity gaps and communication outcomes; however, the results have been inconsistent and from only one relational party. For example, Kam and Hecht (2009) observed associations between the personal–enacted identity gap and topic avoidance, communication satisfaction, and relationship satisfaction, whereas the personal–relational gap failed to reach statistical significance. Jung and Hecht (2008) suggested that there is a close relationship between identity gaps and communication outcomes, and the results of this study suggest that the relationship may be best tested by modeling dyadic data. More specifically, the significant partner effects for GPs’ identity gaps in relation to GC's communication satisfaction foregrounds the notion that perceptions of a relational partner’s identity frames is related to communication. Given recent research that highlights the mutual influence GPs and GC have on each other (Matos & Neves, 2012), our results bring to light a significant theoretical advancement for CTI.

Implications for caregiving with respect to GPs and GC

Overall, this study’s results also have serious implications for GP-GC caregiving. With respect to actor effects, this study demonstrated how GPs’ and GC's identities are related to the quality of their communication, which is related to GPs’ perceptions of their GC's future caregiving intentions, along with GC's actual intention to care for their GP. Although intentions do not guarantee that GC will provide care in the future, the theory of planned behavior (Ajzen, 1991) suggests that individuals’ intentions predict how likely they are to actually care for their GP in the future. Feeling confident that one’s family member will provide assistance when needed is likely to create a sense of security for GPs, which may also enhance GPs’ communication and relationship with their GC. Given the growing number of older adults and the increased need for family caregivers (Fruhauf et al., 2006), this study provides insight into GC’s potential to contribute to the caregiving process.

Past research on identity gaps in the context of the GP-GC relationship has focused on how those gaps reflected at the individual level, but what individuals may think is internalized can actually be detected by a relational partner. For example, as GPs in our sample experienced any of the two identity gaps, GC were less likely to be satisfied with their communication and, in turn, GC reported weaker caregiving intentions. Thus, our findings have practical implications for programs about family caregiving more broadly. In terms of actor effects, programs may focus on (a) helping GPs to reconcile their identity gaps and (b) providing GPs and GC with strategies to enhance communication satisfaction. Past literature suggests that communication particularly dissatisfying when one generation feels that the other is not communicatively attuned to their needs (Williams & Giles, 1996). Thus, providing GPs and GC with advice on how to navigate intergenerational communication, such as how to effectively provide communication-based emotional support (Mansson et al., 2010), may help to increase communication satisfaction and thereby GC's caregiving intentions and GPs’ perceptions of those intentions.

Second, despite partner effects not being able to explain GPs’ perceptions of caregiving intensions, it is important to realize that communication is dynamic, and perceptions that GC will “be there” for the GP may strengthen the GP-GC relationship (Kam & Nussbaum, 2008). For example, one explanation for the reported partner effects of identity gaps on communication satisfaction is that identity gaps create a negative communicative environment (e.g., decreased affectionate communication) within the GP-GC relationship, which may in turn be negatively related to GC's future caregiving intentions. Obtaining a better understanding of the family communication environment and the nature of the GP-GC relationship may shed light on how to circumvent potentially negative health consequences for GPs and increase the quantity and quality of care by GC in the future.

Limitations and future research

Despite this study’s findings, it has several limitations. In particular, the sample of undergraduate students and their GPs may not be representative of other GP-GC dyads outside of the university context (Barker, 2007). Similarly, GPs and young adult GC were mainly European American (93.5%) and may not generalize to other ethnic groups. Indeed, Soliz, Thorson, and Rittenour (2009) suggested that multiracial and multiethnic families might particularly struggle with communicating identity. Furthermore, young adult GC’s choice of GP was limited to a living biological GP who had Internet access. This resulted in a majority reporting on grandmothers (68% of the sample). Harwood (2004) has emphasized the importance of studying GPs’ use of the Internet to express aspects of their identity. Nevertheless, although 6 in 10 adults aged 65 and older now have access to the Internet, some may lack the means to efficiently use the Internet (Smith, 2014). As such, access to and skills to use the Internet may have excluded a number of GPs from participation. Taken together, these limitations suggest that future research should include a more diverse sample to assess the applicability of this study’s findings.

A final limitation concerns the exclusion of family identity. Shared family identity has been related to more positive perceptions of older adults (Soliz & Harwood, 2006) and more supportive communication (Soliz, 2007). A sense of shared family identity may lessen the effects of individual identity gaps on communication satisfaction and provide a highly valued common familial identity on which to base their communication (Soliz, Thorson, & Rittenour, 2009). Future studies of GP and GC identity gaps on communication satisfaction should consider the moderating role of shared family identity.

This study also has several methodological limitations. First, mediation was examined using cross-sectional data. Theoretically, mediation means that X has an effect on Y through one or more intervening variables, M (Preacher & Hayes, 2008), thereby assuming specific time ordering of the variables. Longitudinal data allow for a more accurate representation of mediation compared to cross-sectional data. Second, although GPs and young adult GC were asked to independently complete the survey in the privacy of their own homes, it is possible that participants did not follow these instructions. Third, the self-reported data represent perceptions rather than actual caregiving intentions; because of social desirability, GPs and young adult GC may have reported higher communication satisfaction and caregiving intentions.

Regardless of this study’s limitations, the results reveal how identity gaps are directly and indirectly related to communication quality and, in turn, GC's caregiving intentions. Although research often focuses on parents as primary caregivers to their aging parents, GC also play an integral role in contributing to the care of their GP(s) as primary or secondary caregivers (Even-Zohar & Sharlin, 2009). By modeling dyadic data, this study demonstrated the interdependent nature of GP-GC identity gaps, communication quality, and perceptions of caregiving intentions and their applicability to an understudied group of older adults.

Footnotes

Authors’ note

Portions of this manuscript have been presented previously at the annual convention of the National Communication Association, November 2014, Chicago, Illinois, USA.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

References

Ajzen

(1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50, 179–211.

Anderson

Harwood

Hummert

M. L

(2006). The grandparent–grandchild relationship. Human Communication Research, 31, 268–294. doi:10.1111/j.1468-2958.2005.tb00872.x

Arnett

J. J.

(2000). Emerging adulthood: A theory of development from the late teens through the twenties. American Psychologist, 55, 469–480. doi: 10.1037/0003-066X.55.5.469

Barker

(2007). Young adults’ reactions to grandparent painful self-disclosure: The influence of grandparent sex and overall motivations for communication. International Journal of Aging and Human Development, 64, 195–215. doi: 10.2190/KTNU-0373-20W7-4781

Beaudoin

C. E.

Thorson

(2006). The social capital of Blacks and Whites: Differing effects of the mass media in the United States. Human Communication Research, 32, 157–177.

Browne

M. W.

Cudeck

(1993). Alternative ways of assessing model fit. In Bollen

K. A.

Long

J. S.

(Eds.), Testing structural equation models (pp. 136–161). Newbury Park, CA: Sage.

Brussoni

M. J.

Boon

S. D.

(1998). Grandparental impact in young adults’ relationships with their closest grandparents: The role of relationship strength and emotional closeness. The International Journal of Aging and Human Development, 46, 267–286. doi:10.2190/2B7B-E3LH-74JJ-UJRY

Dellmann-Jenkins

Blankemeyer

Pinkard

(2000). Young adult children and grandchildren in primary caregiver roles to older relatives and their service needs. Family Relations, 49, 177–186. doi: 10.1111/j.1741-3729.2000.00177.x

Drummond

D. K.

Orbe

M. P.

(2009). “Who are you trying to be?”: Identity gaps within intraracial encounters. Qualitative Research Reports in Communication, 10, 81–87. doi: 10.1080/17459430903236098

10.

Even-Zohar

Sharlin

(2009). Grandchild: Adult grandchildren’s perception of their role towards their grandparents from an intergenerational perspective. Journal of Comparative Family Studies, 40, 167–185.

11.

Faulkner

S. L.

Hecht

M. L.

(2011). The negotiation of closetable identities: A narrative analysis of lesbian, gay, bisexual, transgendered queer Jewish identity. Journal of Social and Personal Relationships, 28, 829–847. doi:10.1177/0265407510391338

12.

Fowler

Soliz

(2010). Responses of young adult grandchildren to grandparents’ painful self-disclosures. Journal of Language and Social Psychology, 29, 75–100. doi:10.1177/0261927X09351680

13.

Fruhauf

C. A.

Jarrott

S. E.

Allen

K. R.

(2006). Grandchildren’s perceptions of caring for grandparents. Journal of Family Issues, 27, 887–911. doi:10.1177/0192513X05286019

14.

Geurts

Van Tilburg

T.G.

Poortman

(2012). The grandparent-grandchild relationship in childhood and adulthood: A matter of continuation. Personal Relationships, 19, 267–278. doi: 10.1111/j.1475-6811.2011.01354.x

15.

Hamill

S. B.

(2012). Caring for grandparents with Alzheimer’s disease: Help from the “forgotten” generation. Journal of Family Issues, 33, 1195–1217. doi:10.1177/0192513X12444858

16.

Hamon

R. R.

(1988). Filial responsibility expectations among adult child-older parent pairs. Paper presented at the 41st Annual Scientific Meeting of the Gerontology Society of American, San Francisco.

17.

Harwood

(2004). Relational, role, and social identity as expressed in grandparents’ personal websites. Communication Studies, 55, 300–318. doi: 10.1080/10510970409388621

18.

Harwood

Lin

M.-C.

(2000). Affiliation, pride, exchange, and distance in grandparents’ accounts of relationships with their college-aged grandchildren. Journal of Communication, 50, 31–47. doi: 10.1111/j.1460-2466.2000.tb02851.x

19.

Hecht

M. L.

(1978). Measure of communication satisfaction. Human Communication Research, 4, 350–368. doi: 10.1111/j.1468-2958.1978.tb00721.x

20.

Hecht

M. L.

(1993). 2002 – A research odyssey: Toward the development of a communication theory of identity. Communication Monographs, 60, 76–81. doi: 10.1080/03637759309376297

21.

Hecht

M. L.

Faulkner

S. L.

(2000). Sometimes Jewish, sometimes not: The closeting of Jewish American identity. Communication Studies, 51, 372–387.

22.

Hecht

M. L.

Warren

Jung

Krieger

(2005). The communication theory of identity: Development, theoretical perspective, and future directions. In Gudykunst

W. B.

(Ed.), Theorizing about intercultural communication (pp. 257–278). Newbury Park, CA: Sage.

23.

Heru

A. M.

Ryan

C. E.

Iqbal

(2004). Family functioning in the caregivers of patients with dementia. International Journal of Geriatric Psychology, 19, 533–537. doi: 10.1002/gps.1119

24.

Hillcoat-Nallétamby

Phillips

J. E.

(2011). Sociological ambivalence revisited. Sociology, 45, 202–217.

25.

Bentler

P. M.

(1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6, 1–55. doi: 10.1080/10705519909540118

26.

Jung

(2013). Delineation of a threefold relationship among communication input variables, identity gaps, and depressive symptoms. Southern Communication Journal, 78, 163–184. doi:10.1080/1041794X.2012.741652

27.

Jung

Hecht

M. L.

(2004). Elaborating the communication theory of identity: Identity gaps and communication outcomes. Communication Quarterly, 52, 265–283. doi: 10.1080/01463370409370197

28.

Jung

Hecht

M. L.

(2008). Identity gaps and level of depression among Korean immigrants. Health Communication, 23, 313–325. doi: 10.1080/10410230802229688

29.

Jung

Hecht

M. L

Wadsworth

B. C.

(2007). The role of identity in international students’ psychological well-being in the United States: A model of depression level, identity gaps, discrimination, and acculturation. International Journal of Intercultural Relations, 31, 605–624.

30.

Kam

J. A.

Hecht

M. L.

(2009). Investigating the role of identity gaps among communicative and relational outcomes within the grandparent-grandchild relationship: The young-adult grandchildren’s perspective. Western Journal of Communication, 73, 456–480. doi: 10.1080/10570310903279067

31.

Kam

J. A.

Nussbaum

(2008, 11). Exploring the dynamic nature of the grandparent-grandchild relationship. Paper presented at the annual meeting of the National Communication Association, San Diego, CA.

32.

Kemp

C. L.

(2007). Grandparent-grandchild ties: Reflections on continuity and change across three generations. Journal of Family Issues, 28, 855–881. doi:10.1177/0192513X07299599

33.

Kenny

D. A.

Kashy

D. A.

Cook

W. L.

(2006). Dyadic data analysis. New York, NY: The Guilford Press.

34.

Kline

R. B.

(2005). Principles and practice of structural equation modeling. New York, NY: The Guilford Press.

35.

Lin

M-C.

Harwood

(2003). Predictors of grandparent-grandchild relational solidarity in Taiwan. Journal of Social and Personal Relationships, 20, 537–563. doi:10.1177/02654075030204006

36.

Mansson

D. H.

Myers

S. A.

Turner

L. H.

(2010). Relational maintenance behaviors in the grandchild-grandparent relationship. Communication Research Reports, 27, 68–79. doi: 10.1080/08824090903526521

37.

Matos

A. D.

Neves

R. B.

(2012). Understanding adolescent grandchildren’s influence on their grandparents. In Arber

Timonen

(Eds.), Contemporary grandparenting (pp. 203–225). Great Britain: The Policy Press.

38.

Merrill

D. M.

(1996). Conflict and cooperation among adult siblings during the transition to the role of filial caregiver. Journal of Social and Personal Relationships, 13, 399–413. doi:10.1177/0265407596133006

39.

Muthén

L. K.

Muthén

B. O.

(1998–2007). Mplus user’s guide. (5th ed.). Los Angeles, CA: Muthén & Muthén.

40.

Nussbaum

J. F.

Bettini

(1994). Shared stories of the grandparent-grandchild relationship. International Journal of Aging and Human Development, 39, 67–80. doi:10.2190/7WPK-LM6C-QCA4-GQ4R

41.

Piercy

K. W.

(2007). Characteristics of strong commitments to intergenerational family care of older adults. Journal of Gerontology: Social Sciences, 62B, S381–S387.

42.

Piercy

K. W.

Chapman

J. G.

(2001). Adopting the caregiver role: A family legacy. Family Relations, 50, 386–393. doi: 10.1111/j.1741-3729.2001.00386.x

43.

Preacher

K. J.

Hayes

A. F.

(2008). Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior Research Methods, 40, 879–891. doi: 10.3758/BRM.40.3.879

44.

Ramirez

Jr (2008). An examination of the tripartite approach to commitment: An actor-partner interdependence model analysis of the effect of relational maintenance behavior. Journal of Social and Personal Relationships, 25, 943–965. doi:10.1177/0265407508100309

45.

Ross

M. E. T.

Aday

L. A.

(2006). Stress and coping in African American grandparents who are raising their grandchildren. Journal of Family Issues, 27, 912–932. doi:10.1177/0192513X06287167

46.

Slater

M. D.

Hayes

A. F.

Ford

V. L.

(2007). Examining the moderating and mediating roles of news exposure and attention on adolescent judgments of alcohol-related risks. Communication Research, 34, 355–381. doi:10.1177/0093650207302783

47.

Smith

(2014, April). Older adults and technology use. Washington, DC: Pew Internet & American Life Project.

48.

Soliz

(2007). Communicative predictors of a shared family identity: Comparison of grandchildren’s perceptions of family-of-origin grandparents and stepgrandparents. Journal of Family Communication, 7, 177–194. doi:10.1080/15267430701221636

49.

Soliz

Harwood

(2006). Shared family identity, age salience, and intergroup contact: Investigation of the grandparent-grandchild relationship. Communication Monographs, 73, 87–107. doi: 10.1080/03637750500534388

50.

Soliz

J. E.

Lin

M. -C.

Anderson

Harwood

(2006). Friends and allies: Communication in grandparent-grandchild relationships. In Floyd

Morman

M. T.

(Eds.), Widening the family circle: New research on family communication (pp. 65–79). Thousand Oaks, CA: Sage.

51.

Soliz

Thorson

A. R.

Rittenour

C. E.

(2009). Communicative correlates of satisfaction, family identity, and group salience in multiracial/ethnic families. Journal of Marriage and Family, 71, 819–832. doi:10.1111/j.1741-3737.2009.00637.x

52.

Spillman

B. C.

Pezzin

L. E.

(2000). Potential and active family caregivers: Changing networks and the “sandwich generation.” The Milbank Quarterly, 78, 347–374. doi: 10.1111/1468-0009.00177

53.

Villar

Triado

Pinazo-Hernandis

Celdran

Solé

(2010). Grandparents and their adolescent grandchildren: Generational stake or generational complaint? Journal of Intergenerational Relationships, 8, 281–297. doi:10.1080/15350770.2010.498759

54.

Vincent

G. K.

Velkoff

V. A.

(2010). The next four decades: The older population in the United States 2010 to 2050. Current Population Reports. Washington, DC: U.S. Census Bureau, pp. 1–14.

55.

Wadsworth

B. C.

Hecht

M. L.

Jung

(2008). The role of identity gaps, discrimination, and acculturation in international students’ educational satisfaction in American classrooms. Communication Education, 57, 64–87. doi: 10.1080/03634520701668407

56.

Whitechurch

G. G.

Constantine

L. L.

(1993). Systems theory. In Boss

P. G.

Doherty

W. J.

LaRossa

Schumm

W. R.

Steinmetz

S. K.

(Eds.), Sourcebook of family theories and methods (pp. 325–355). New York, NY: Springer.

57.

Williams

Giles

(1996). Intergenerational conversations: Young adults’ retrospective accounts. Human Communication Research, 23, 220–250. doi: 10.1111/j.1468-2958.1996.tb00393.x

58.

Williams

Nussbaum

J. F.

(2001). Intergenerational communication across the life span. Mahwah, NJ: Erlbaum.

59.

Wolff

J. L.

Kasper

J. D.

(2006). Caregivers of frail elders: Updating a national profile. The Gerontologist, 46, 344–356. doi:10.1093/geront/46.3.344